Abstract
Van't Hoff's work constitutes the first systematic contribution to experimental petrology. At all times, the problem was perceived as geologic in nature and the laboratory results were checked against natural assemblages whenever possible. The phase rule was not used, nor, for that matter, was chemical thermodynamics, except for the Van't Hoff equation. However, the work of Van't Hoff and Van Deventer was indirectly involved in the evolution of phase theory by Roozeboom, Van Rijn van Alkemade, and Schreinemakers. Meyerhoffer himself wrote the first text explicitly devoted to the phase rule. The impact of Van't Hoff's study was enormous, but it was restricted to those geologists willing and able to cope with chemistry. Foremost among them were igneous petrologists who had long since accepted chemical arguments for classification purposes. I consider the Geophysical Laboratory program to be the most direct heir of the Van't Hoff approach. Although the shape of that program was formulated independently by Van Hise, Becker, Day, and others, the inspiration they derived from Van't Hoff's successes is clearly acknowledged. The study of the fusion of plagioclases by Day and Allen (41), which directly led to the authorization for the Geophysical Laboratory, was the igneous counterpart of Van't Hoff's low-temperature experimental petrology. On metamorphic petrology, too, Van't Hoff left his mark, with V. M. Goldschmidt acting as his disciple. The interpretation of the Kristiania contact rocks was explicitly based on Van't Hoff's double salt law in preference to the phase rule. Sedimentologists remained unaffected and continued their preoccupation with description and classification. Chemical arguments remained subordinate in their work and of an elementary nature, underscoring the chasm between "hard" rocks and "soft" rocks. This gulf is only now beginning to close as a result of the blossoming of experimental petrology and geochemistry since World War II. At last the generality of the point of view of Gibbs is being accepted. If Van't Hoff's contribution had been appreciated fully at the time, this could have happened 70 years earlier.

This publication has 9 references indexed in Scilit: